Resource allocation method based on weighted chaotic bat algorithm in ultra-dense networks

Abstract

The invention discloses a resource allocation method based on a weighted chaotic bat algorithm in an ultra-dense network. Firstly, the traditional bat algorithm is mainly used to solve the nonlinear problem of continuous solution, but it is a discrete solution for resource block allocation, so the original algorithm needs to be modified to find the optimal solution for the discrete solution problem; in addition, the invention also uses the ergodicity and randomness of chaotic mapping to generate a chaotic sequence to replace the uniformly distributed random number in the original bat algorithm, thereby increasing the diversity of the population; in addition, the logarithmic decreasing inertia weight is used to accelerate the convergence speed; finally, the invention not only considers the situation that the number of resource blocks is greater than the number of home base station user equipment, but also considers the situation that under the state of resource scarcity, that is, when the number of resource blocks is less than the number of home base station user equipment, different initial population encoding methods are used, so as to make full use of resources and improve the robustness of the algorithm; in general, the method can effectively solve the resource allocation problem in the ultra-dense network and improve the throughput of the system by optimizing the bat algorithm.

Description

Technical Field

[0001] The invention belongs to the field of wireless communications, and in particular relates to a resource allocation method based on a weighted chaotic bat algorithm in an ultra-dense network, and is mainly applied to resource block allocation of users in the ultra-dense network. Background Art

[0002] With the rapid development of wireless mobile networks, a large number of smart devices are emerging. In order to meet the surging traffic demand, ultra-dense networks will be one of the key technologies of 5G. In ultra-dense networks, in order to ensure indoor wireless coverage, especially in areas with dense buildings, simply increasing the transmission power of macro base stations cannot solve the problem. The dense deployment of low-power home base stations can not only enhance cellular coverage and alleviate macro cellular traffic in high-density areas, but also improve network throughput and spectrum efficiency. This is widely considered to be a promising technology. However, since densely deployed home base stations will bring huge computational workloads and interference problems, the use of semi-distributed resource management and allocation has become increasingly popular. For semi-distributed resource management and allocation, the families are first clustered, and then the cluster heads are used in each cluster to reasonably allocate resources. As a new type of swarm intelligence optimization algorithm, the bat algorithm has a strong optimization ability, but the algorithm mainly targets the problem of continuous solutions, so the traditional bat algorithm cannot directly solve the problem of resource block allocation. In addition, the algorithm faces the problem of premature maturity and the possibility of falling into a local optimal solution.

[0003] In response to the above technical problems, the system throughput can be effectively increased by improving the bat algorithm. Summary of the invention

[0004] The purpose of the present invention is to provide a resource allocation method based on a weighted chaotic bat algorithm in an ultra-dense network to address the defects of the prior art.

[0005] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A resource allocation method based on weighted chaotic bat algorithm in an ultra-dense network comprises the following steps:

[0007] S1, building a channel allocation model;

[0008] S2, initializing the relevant parameters of the weighted chaotic bat algorithm;

[0009] S3, calculate the fitness based on the initial bat population and find the current optimal bat position;

[0010] S4, calculating the logarithmic decreasing inertia weight;

[0011] S5, update the sound wave frequency, speed and position;

[0012] S6, generate a random number between 0 and 1, and determine whether the random number is greater than the current frequency. If so, a new local solution is generated, otherwise a new solution is generated by random flight;

[0013] S7, generate a random number between 0 and 1, determine whether the random number is less than the current loudness and the new solution is better than the current solution, if both conditions are met, accept the new solution and update the sound wave loudness and frequency, then detect whether the new solution is better than the current optimal solution, and update the number of iterations;

[0014] S8. Determine whether the maximum number of iterations has been reached. If so, output the resource allocation result. Otherwise, repeat steps S4 to S7.

[0015] Furthermore, in step S1, the channel allocation model is expressed as:

[0016]

[0017] Where R k represents the throughput of the kth cluster, k, l, and i represent the cluster number, resource block number, and home base station user number, respectively; K represents the total number of ultra-dense network clusters, L represents the total number of resource blocks, and J k represents the number of users in cluster k; Represents the resource block allocation status, represents the allocation of resource block l to the i-th home base station user in cluster k, represents that resource block l is not allocated to the i-th home base station user in cluster k; ΔB represents the bandwidth of the subchannel, represents the transmit power of the ith base station in cluster k on resource block l, represents the channel gain from base station f in cluster n to user i in cluster k on resource block l, where it is assumed that one home base station serves one user; represents the sum of interference within the cluster, represents the sum of interference between clusters; δ 2 represents the Gaussian additive white Gaussian noise method; C1 represents the resource block allocation status, C2 indicates that multiple resource blocks are allowed to be allocated to one user. In order to ensure full utilization of resources, after each user is allocated a resource block, the excess subchannels can be allocated to existing users. At the same time, when resources are scarce, that is, the number of users is greater than the number of resource blocks, it is guaranteed that the user can be allocated at least one resource block; in C3 represents the throughput achieved by user i in cluster k. If it is greater than zero, it means that each base station can have a certain throughput, ensuring that the base station is allocated at least one resource block. As can be seen from the equation, maximizing the throughput within the cluster through resource block allocation can make the throughput of the entire system close to maximum.

[0018] Furthermore, in step S2, the content of the weighted chaotic bat algorithm initialization includes the bat population number U; the upper limit of the number of iterations T max ; Upper limit of sound wave frequency f max , lower limit of sound wave frequency f min ; Sound wave attenuation coefficient α; Frequency increase coefficient γ; Bat speed v 0 , 0 represents the initialized value (data of the 0th generation); bat population Represents the initialization of U bats; using Gauss mapping to generate a chaotic sequence as the initial value of the sound wave loudness Using Tent mapping to generate chaotic sequence as frequency initial value

[0019]

[0020] When the number of users is greater than or equal to the number of resource blocks, the initial bat position is expressed as:

[0021]

[0022] in The dimension is J k , represents the initial position of the u-th bat. The chaotic sequence is used to generate random distribution results during initialization, and the position of each bat represents a resource block allocation result, each allocation result represents a solution, that is, the dimension of the solution is J k . represents the resource block number allocated to the i-th user of the u-th bat during initialization, round(*) represents rounding *, rand represents a random number between 0 and 1, which follows a uniform distribution. And they are all integers. Each bat position represents a resource block allocation result. It means that the i-th user of the u-th bat is allocated the l-th resource block during initialization. Here, Gauss mapping is used instead of rand to generate random numbers.

[0023] When the number of users is less than the number of resource blocks, the initial bat position is expressed as:

[0024]

[0025] in The dimension of is L, that is, the dimension of the solution is L, Represents the initial position of the u-th bat. represents the user number to which the l-th resource block of the u-th bat is allocated during initialization, And they are all integers. It means that during initialization, the resource block l is allocated to user i in the u-th bat. Here, Gauss mapping is used instead of rand to generate random numbers.

[0026] Further, the Gauss mapping is expressed as:

[0027]

[0028] The tent mapping is expressed as:

[0029]

[0030] Here, we mainly use the ergodicity and randomness of chaotic mapping to generate chaotic sequences to replace the uniformly distributed random numbers of the original bat algorithm. Whether it is Gauss or Tent mapping, the χ d Both represent the current (dth) value, χ d+1 Represents the value generated next time, the initial chaotic map is a random number between 0 and 1.

[0031] Furthermore, in step S3, the fitness is the total throughput achieved by the users in the cluster, and the current optimal bat position is the allocation method that maximizes the user throughput in the existing allocation results.

[0032] Furthermore, in step S4, the logarithmically decreasing inertia weight is expressed as:

[0033] ω=ω max +(ω min +ω max )×log 10 (a+10t / T max )

[0034] Where ω represents the logarithmically decreasing inertia weight value; ω min ,ω max represents the adjustment coefficient of the inertia weight value; a is a constant used to adjust the update speed; t represents the current number of iterations; T max Represents the upper limit of the number of iterations.

[0035] Further, in step S5, the sound wave frequency is updated as follows:

[0036]

[0037] in Represents the sound wave frequency of the u-th bat in the t-th generation, where the value of the Gaussian map is used instead of rand.

[0038] The speed update is expressed as:

[0039]

[0040] in represents the speed of the u-th bat in the t-th generation; ω represents the logarithmically decreasing inertia weight value; x * Represents the current optimal solution; represents the position of the u-th bat in generation t-1.

[0041] The location update is expressed as:

[0042]

[0043] in represents the position of the u-th bat in the t-th generation.

[0044] Furthermore, in step S6, the generation of the local new solution is expressed as:

[0045]

[0046] where ε∈[-1,1] is a uniformly distributed random number; x old represents the previous solution, x new Represents the new solution generated; represents the average loudness of the tth generation in the entire population.

[0047] Furthermore, in step S7, the sound wave loudness update is expressed as:

[0048]

[0049] in Represents the sound wave loudness of the u-th bat in the t-th generation. When the bat approaches its prey, it will reduce the sound wave loudness; 0≤α≤1, represents the sound wave loudness attenuation coefficient.

[0050] The sound wave frequency update is expressed as:

[0051]

[0052] in represents the sound wave frequency of the u-th bat in the t-th generation; 0≤γ≤1, represents the frequency increase coefficient. As the bats continue to approach the optimal solution, the sound wave frequency tends to the initial sound wave frequency

[0053] This invention proposes a resource allocation method based on a weighted chaotic bat algorithm in an ultra-dense network. First of all, the traditional bat algorithm is mainly used to solve the nonlinear problem of continuous solutions, but it is a discrete solution for resource block allocation. Therefore, the original algorithm needs to be modified to be suitable for the problem of discrete solutions to find the best. Not only that, the present invention also uses the ergodicity and randomness of chaotic mapping to generate a chaotic sequence to replace the uniformly distributed random numbers of the original bat algorithm, thereby increasing the diversity of the population. In addition, the logarithmic decreasing inertia weight is used to accelerate the convergence speed. Finally, the present invention not only considers the situation where the number of resource blocks is greater than or equal to the number of users, but also considers the situation where the number of resource blocks is greater than the number of users in the cluster, that is, when the number of resource blocks is less than the number of users in the cluster, using different initial population encoding methods can make full use of resources and improve the robustness of the algorithm. In general, this method can effectively solve the resource allocation problem in ultra-dense networks and improve the throughput of the system by optimizing the bat algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is the algorithm flow chart of the embodiment of the present invention

[0055] Figure 2 This is a CDF comparison diagram of the system spectrum efficiency of an embodiment of the present invention.

[0056] Figure 3 This is a comparison diagram of the relationship between the number of resource blocks and the average user throughput in an embodiment of the present invention. Specific implementation methods

[0057] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can better understand the scheme of the present invention from the following content.

[0058] The present invention first simulates the distribution position of home base stations in an ultra-dense network on a 400m*400m two-dimensional plane, where the position of the home base station is randomly generated, and users are randomly distributed around the base station. The coverage radius of the home base station is generally 10m. Here, in order to be more in line with the actual application scenario, not only the maximum distance between the base station and the user is set to 10m, but also the minimum distance is set to 2m. Assume that all home base stations are currently widely used closed access (Closed Subscriber Group, CSG) configurations, each base station only serves one user, and the resource blocks are mutually orthogonal. Here we have used the existing clustering algorithm to reasonably cluster them. In this example, the system is divided using the K-means algorithm using the location information of the base station. Then the present invention mainly uses the cluster head in each cluster to allocate resource blocks for the system that has been clustered. The channel model is mainly composed of path loss, wall penetration loss and shadow fading. The path loss PL = 38.46 + 20logd, and only the wall penetration loss of the interference signal is considered. The specific system simulation parameters are shown in Table 1:

[0059] Table 1 System simulation parameters

[0060]

[0061] like Figure 1 As shown, the specific process of the resource allocation method based on the weighted chaotic bat algorithm in the ultra-dense network of the present invention comprises the following steps:

[0062] S1. Construct channel allocation model;

[0063] S12. The channel allocation model is expressed as:

[0064]

[0065] Where R k represents the throughput of the kth cluster, k, l, and i represent the cluster number, resource block number, and home base station user number, respectively; K represents the total number of ultra-dense network clusters, L represents the total number of resource blocks, and J k represents the number of users in cluster k; Represents the resource block allocation status, represents the allocation of resource block l to the i-th home base station user in cluster k, represents that resource block l is not allocated to the i-th home base station user in cluster k; ΔB represents the bandwidth of the subchannel, represents the transmit power of the ith base station in cluster k on resource block l, represents the channel gain from base station f in cluster n to user i in cluster k on resource block l, where it is assumed that one home base station serves one user; represents the sum of interference within the cluster, represents the sum of interference between clusters; δ 2 represents the Gaussian additive white Gaussian noise method; C1 represents the resource block allocation status, C2 indicates that multiple resource blocks are allowed to be allocated to one user. In order to ensure full utilization of resources, after each user is allocated a resource block, the excess subchannels can be allocated to existing users. At the same time, when resources are scarce, that is, the number of users is greater than the number of resource blocks, it is guaranteed that the user can be allocated at least one resource block; in C3 represents the throughput achieved by user i in cluster k. If it is greater than zero, it means that each base station can have a certain throughput, ensuring that the base station is allocated at least one resource block. As can be seen from the equation, maximizing the throughput within the cluster through resource block allocation can make the throughput of the entire system close to maximum.

[0066] S2. Initialize the relevant parameters of weighted chaotic bat algorithm;

[0067] S21. According to the resource allocation method based on the weighted chaotic bat algorithm in the ultra-dense network, it is characterized in that in step S2, the content of the weighted chaotic bat algorithm initialization includes the bat population number U; the upper limit of the number of iterations T max ; Upper limit of sound wave frequency f max , lower limit of sound wave frequency f min ; Sound wave attenuation coefficient α; Frequency increase coefficient γ; Bat speed v 0 , 0 represents the initialized value (data of the 0th generation); bat population Represents the initialization of U bats; using Gauss mapping to generate a chaotic sequence as the initial value of the sound wave loudness Using Tent mapping to generate chaotic sequence as frequency initial value

[0068] When the number of users is greater than or equal to the number of resource blocks, the initial bat position is expressed as:

[0069]

[0070] in The dimension is J k , represents the initial position of the u-th bat. The chaotic sequence is used to generate random distribution results during initialization, and the position of each bat represents a resource block allocation result, each allocation result represents a solution, that is, the dimension of the solution is J k . represents the resource block number allocated to the i-th user of the u-th bat during initialization, round(*) represents rounding *, rand represents a random number between 0 and 1, which follows a uniform distribution. And they are all integers. Each bat position represents a resource block allocation result. It means that the i-th user of the u-th bat is allocated the l-th resource block during initialization. Here, Gauss mapping is used instead of rand to generate random numbers.

[0071] When the number of users is less than the number of resource blocks, the initial bat position is expressed as:

[0072]

[0073] in The dimension of is L, that is, the dimension of the solution is L, Represents the initial position of the u-th bat. represents the user number to which the l-th resource block of the u-th bat is allocated during initialization, And they are all integers. It means that during initialization, the resource block l is allocated to user i in the u-th bat. Here, Gauss mapping is used instead of rand to generate random numbers.

[0074] The initialization parameter settings related to the weighted chaotic bat algorithm are shown in Table 2:

[0075] Table 2 Parameters of weighted chaotic bat algorithm

[0076]

[0077] S22. Here, the ergodicity and randomness of the chaotic mapping are mainly used to generate a chaotic sequence to replace the uniformly distributed random numbers of the original bat algorithm. The Gauss mapping is expressed as:

[0078]

[0079] The tent mapping is expressed as:

[0080]

[0081] Whether it is Gauss or Tent mapping, χ in the equation d Both represent the current (dth) value, χ d+1 Represents the value generated next time, the initial chaotic map is a random number between 0 and 1.

[0082] S3. Calculate the fitness based on the initial bat population and find the current optimal bat position;

[0083] S31. Fitness is the total throughput achieved by users in the cluster. The current optimal bat position is the allocation method that maximizes the user throughput among the existing allocation results.

[0084] S4. Calculate the logarithmic decreasing inertia weight;

[0085] S41. According to the resource allocation method based on weighted chaotic bat algorithm in the ultra-dense network, it is characterized in that in step S4, the logarithmic decreasing inertia weight is expressed as:

[0086] ω=ω max +(ω min +ω max )×log 10 (a+10t / T max )

[0087] Where ω represents the logarithmically decreasing inertia weight value; ω min ,ω max represents the adjustment coefficient of the inertia weight value; a is a constant used to adjust the update speed, in this example a = 1; t represents the current number of iterations; T maxRepresents the upper limit of the number of iterations.

[0088] S5. Update the sound wave frequency, speed and position;

[0089] S51. According to the resource allocation method based on weighted chaotic bat algorithm in the ultra-dense network, it is characterized in that in step S5, the sound wave frequency update is expressed as:

[0090]

[0091] in Represents the sound wave frequency of the u-th bat in the t-th generation, where the value of the Gaussian map is used instead of rand.

[0092] The speed update is expressed as:

[0093]

[0094] in represents the speed of the u-th bat in the t-th generation; ω represents the logarithmically decreasing inertia weight value; x * Represents the current optimal solution; represents the position of the u-th bat in generation t-1.

[0095] The location update is expressed as:

[0096]

[0097] in represents the position of the u-th bat in the t-th generation.

[0098] S6. Generate a random number between 0 and 1, and determine whether the random number is greater than the current frequency. If so, a new local solution is generated, otherwise a new solution is generated by random flight;

[0099] S61. The generation of the local new solution is expressed as:

[0100]

[0101] where ε∈[-1,1] is a uniformly distributed random number; x old represents the previous solution, x new Represents the new solution generated; represents the average loudness of the tth generation in the entire population.

[0102] S7. Generate a random number between 0 and 1, and determine whether the random number is less than the current loudness and the new solution is better than the current solution. If both conditions are met, accept the new solution and update the sound wave loudness and frequency, then detect whether the new solution is better than the current optimal solution and update the number of iterations;

[0103] S71. According to the resource allocation method based on weighted chaotic bat algorithm in the ultra-dense network, it is characterized in that in step S7, the sound wave loudness update is expressed as:

[0104]

[0105] in Represents the sound wave loudness of the u-th bat in the t-th generation. When the bat approaches its prey, it will reduce the sound wave loudness; 0≤α≤1, represents the sound wave loudness attenuation coefficient.

[0106] The sound wave frequency update is expressed as:

[0107]

[0108] in represents the sound wave frequency of the u-th bat in the t-th generation; 0≤γ≤1, represents the frequency increase coefficient. As the bats continue to approach the optimal solution, the sound wave frequency tends to the initial sound wave frequency

[0109] S8. Determine whether the maximum number of iterations has been reached. If so, output the resource allocation result. Otherwise, repeat steps S4 to S7.

[0110] Figure 2 This is a comparison diagram of the cumulative distribution function (CDF) of the system spectrum efficiency of the embodiment of the present invention, which describes the effect of the CDF of the system spectrum efficiency in different algorithms. The main comparisons are random channel allocation and greedy algorithm. Here, the number of home base stations is 80 and the number of resource blocks L=9. It can be seen from the effect diagram that the curve of the algorithm of the present invention is steeper, indicating that the system spectrum efficiency is more stable within a certain value range and the algorithm is more stable; while the distribution range of the spectrum efficiency of random channel allocation is obviously wider, mainly because random channel allocation is more random and cannot achieve good stability. In addition, the two algorithms used in the present invention are distributed at a higher spectrum efficiency, which is significantly improved compared with other algorithms. The algorithm in the present invention is superior to other comparison algorithms because the bat algorithm itself can control the global interests, while the greedy algorithm can only enable each user to reach the maximum target value, and when the number of resource blocks is greater than the number of users, it can only be selected from resource blocks that have not been selected. In the end, there will be extra resource blocks that have not been allocated, resulting in resource waste, and it does not have a global view.

[0111] Figure 3This is a comparison chart of the relationship between the number of resource blocks and the average user throughput of the embodiment of the present invention. Here the number of users is 80, and the number of resource blocks L = [7:1:12]. When the number of users is fixed, as the number of resource blocks increases, the average user throughput continues to increase. The reason is that the more resource blocks, the more choices the user has, the more resource blocks that can be used, and the average user throughput will naturally increase. At the same time, it can be seen that the algorithm of the present invention is obviously better than other algorithms. In addition, as the number of resource blocks increases, the gap between the greedy algorithm and the algorithm of the present invention becomes larger and larger, and the growth rate of the greedy algorithm continues to decrease. The main reason is that when the number of resource blocks is greater than the number of users, the greedy algorithm cannot make full use of the excess resource blocks. However, the algorithm designed by the present invention has maintained a high-speed growth because the present invention performs two different encoding methods on the initial population. When the number of user devices is greater than or equal to the number of resource blocks, the present invention adopts the method of user selection of resource blocks for encoding, and each user must make a selection; when the number of users is less than the number of resource blocks, the present invention adopts the method of resource selection of users for encoding. The combination of the two can make the average throughput continue to grow.

[0112] The specific embodiments described above are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A resource allocation method based on weighted chaotic bat algorithm in ultra-dense networks, characterized in that: The following steps are involved: S1, building a channel allocation model; S2, initializing the relevant parameters of the weighted chaotic bat algorithm; S3, calculate the fitness based on the initial bat population and find the current optimal bat position; S4, calculating the logarithmic decreasing inertia weight; S5, update the sound wave frequency, speed and position; S6, generate a random number between 0 and 1, and determine whether the random number is greater than the current frequency. If so, a new local solution is generated, otherwise a new solution is generated by random flight; S7, generate a random number between 0 and 1, determine whether the random number is less than the current loudness and the new solution is better than the current solution, if both conditions are met, accept the new solution and update the sound wave loudness and frequency, then detect whether the new solution is better than the current optimal solution, and update the number of iterations; S8, determine whether the maximum number of iterations has been reached, if so, output the resource allocation result, otherwise repeat steps S4 to S7; In step S1, the channel allocation model is expressed as: Among them, R k represents the throughput of the kth cluster, k, l, and i represent the cluster number, resource block number, and home base station user number, respectively; K represents the total number of ultra-dense network clusters, L represents the total number of resource blocks, and J k represents the number of users in cluster k; Represents the resource block allocation status, represents the allocation of resource block l to the i-th home base station user in cluster k, represents that resource block l is not allocated to the i-th home base station user in cluster k; ΔB represents the bandwidth of the subchannel, represents the transmit power of the ith base station in cluster k on resource block l, represents the channel gain from base station f in cluster n to user i in cluster k on resource block l, where it is assumed that one home base station serves one user; represents the sum of interference within the cluster, represents the sum of interference between clusters; δ 2 represents the Gaussian additive white Gaussian noise method; C1 represents the resource block allocation status, C2 means that multiple resource blocks are allowed to be allocated to one user. When each user is allocated a resource block, the excess subchannels are allocated to the existing users. At the same time, when resources are scarce, that is, the number of users is greater than the resource blocks, it is guaranteed that the user can be allocated at least one resource block; C3 represents the throughput achieved by user i in cluster k. A value greater than zero means that each base station can have a certain throughput, ensuring that the base station is allocated at least one resource block; In step S3, the fitness is the total throughput achieved by the users in the cluster, and the current optimal bat position is the allocation method that maximizes the user throughput in the existing allocation results.

2. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 1 is characterized in that: In step S2, the contents of weighted chaotic bat algorithm initialization include bat population number U; the upper limit of iteration number T max ; Upper limit of sound wave frequency f max , lower limit of sound wave frequency f min ; Sound wave attenuation coefficient α; Frequency increase coefficient γ; Bat speed v 0 , 0 represents the initialization value; bat population Represents that U bats have been initialized; Using Gauss mapping to generate chaotic sequences as initial values ​​of sound wave loudness Using Tent mapping to generate chaotic sequence as frequency initial value When the number of users is greater than or equal to the number of resource blocks, the initial bat position is expressed as: in The dimension is J k , represents the initial position of the u-th bat. The chaotic sequence is used to generate random distribution results during initialization, and the position of each bat represents a resource block allocation result, each allocation result represents a solution, that is, the dimension of the solution is J k ; represents the resource block number allocated to the i-th user of the u-th bat during initialization, round(*) represents rounding *, rand represents a random number between 0 and 1, which follows a uniform distribution. And they are all integers; each bat position represents a resource block allocation result, l∈[1,L] means that the i-th user of the u-th bat is allocated the l-th resource block at the time of initialization. Here, Gauss mapping is used instead of rand to generate random numbers; When the number of users is less than the number of resource blocks, the initial bat position is expressed as: in The dimension of is L, that is, the dimension of the solution is L, represents the initial position of the u-th bat; represents the user number to which the l-th resource block of the u-th bat is allocated during initialization, and are all integers; i∈[1,J k ] represents that during initialization, the resource block l is allocated to user i in the u-th bat. Here, Gauss mapping is used instead of rand to generate random numbers.

3. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 2 is characterized in that: Gauss mapping is expressed as: The tent mapping is expressed as: The chaotic sequence is generated by using the ergodic and random characteristics of chaotic mapping to replace the uniformly distributed random numbers of the original bat algorithm. Whether it is Gauss or Tent mapping, the χ d Both represent the current value, χ d+1 Represents the value generated next time, the initial chaotic map is a random number between 0 and 1.

4. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 1, characterized in that: In step S4, the logarithmic decreasing inertia weight is expressed as: oh = oh max +(ω min +oh max )×log 10 (a+10t / T max ) Where ω represents the logarithmically decreasing inertia weight value; ω min ,ω max The adjustment coefficient representing the inertia weight value; a is a constant used to adjust the update speed; t represents the current iteration number; T max Represents the upper limit of the number of iterations.

5. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 1, characterized in that: In step S5, the sound wave frequency is updated as: in represents the sound wave frequency of the u-th bat in the t-th generation, where the value of the Gaussian map is used instead of rand; The velocity update is expressed as: in represents the speed of the u-th bat in the t-th generation; ω represents the logarithmically decreasing inertia weight value; x * Represents the current optimal solution; represents the position of the u-th bat in generation t-1; The location update is expressed as: in represents the position of the u-th bat in the t-th generation.

6. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 1, characterized in that: In step S6, the generation of the local new solution is expressed as: where ε∈[-1,1] is a uniformly distributed random number; x old represents the previous solution, x new Represents the new solution generated; represents the average loudness of the tth generation in the entire population.

7. The resource allocation method based on weighted chaotic bat algorithm in ultra-dense network according to claim 1 is characterized in that: In step S7, the sound wave loudness update is expressed as: in, represents the sound wave loudness of the u-th bat in the t-th generation. When bats approach their prey, they will reduce the sound wave loudness; 0≤α≤1, represents the sound wave loudness attenuation coefficient; The sound wave frequency update is expressed as: in represents the sound wave frequency of the u-th bat in the t-th generation; 0≤γ≤1, represents the frequency increase coefficient. As the bats continue to approach the optimal solution, the sound wave frequency tends to the initial sound wave frequency

Images